Utils.hs revision 2eb84fc82d3ffa9116bc471fda3742bd9e5a24bb
{-|
> HetCATS/Utils.hs
> $Id$
> Authors: Klaus L�ttich
> Year: 2002
Useful functions that can't be found in the libraries.
But should shared across HetCATS.
Todo:
- Add your own functions.
-}
module Common.Utils where
import Data.List
import Common.AS_Annotation
{- |
A function inspired by perls join function. It joins a list of
lists of elements by seperating them with a seperator element.
-}
joinWith :: a -- ^ seperator element
-> [[a]] -- ^ list of lists of elements
-> [a]
joinWith sep = concat . intersperse (sep:[])
{- |
A function inspired by the perl function split. A list is splitted
on a seperator element in smaller lists. The seperator element is
dropped from the resulting list.
-}
splitOn :: Eq a => a -- ^ seperator
-> [a] -- ^ list to split
-> [[a]]
splitOn sep = (\(f,r) -> f : case r of
[] -> []
_ -> (splitOn sep $ tail r)
) . break ((==) sep)
{-|
A function inspired by a perl function from the standard perl-module
File::Basename. It removes the directory part of a filepath.
-}
basename :: FilePath -> FilePath
basename fp = (\(path,basen) -> basen) (stripDir fp)
{-|
A function inspired by a perl function from the standard perl-module
File::Basename. It gives the directory part of a filepath.
-}
dirname :: FilePath -> FilePath
dirname fp = (\(path,basen) -> path) (stripDir fp)
{-|
A function inspired by a perl function from the standard perl-module
File::Basename. It splits a filepath into the basename, the
directory and gives the suffix that matched from the list of
suffixes. If a suffix matched it is removed from the basename.
-}
fileparse :: [String] -- ^ list of suffixes
-> FilePath
-> (FilePath,FilePath,Maybe String)
-- ^ (basename,directory,matched suffix)
fileparse sufs fp = let (path,base) = stripDir fp
(base',suf) = stripSuffix sufs base
in (base',path,suf)
stripDir :: FilePath -> (FilePath,FilePath)
stripDir fp = (\(x,y) -> (reverse y, reverse x)) (break (== '/') (reverse fp))
stripSuffix :: [String] -> FilePath -> (FilePath,Maybe String)
stripSuffix suf fp = case filter justs $ map (stripSuf fp) suf of
[] -> (fp, Nothing)
((Just (x,s)):_) -> (x,Just s)
where stripSuf f s | s `isSuffixOf` f = Just (stripOf s f, s)
| otherwise = Nothing
justs (Nothing) = False
justs (Just _) = True
stripOf :: (Show a, Eq a) => [a] -> [a] -> [a]
stripOf suf inp = reverse $ stripOf' (reverse suf) (reverse inp)
where stripOf' [] i = i
stripOf' (x:xs) [] = error $
concat ["suffix is longer than input string\n"
,"input was: ", show suf, " ",show inp ]
stripOf' (x:xs) (y:ys) | x == y = stripOf' xs ys
| otherwise =
error $ concat ["suffix don't match input"
," at "
,show $ reverse (x:xs)
," ",show $ reverse (y:ys)]
-- stripOf suf = reverse . drop (length suf) . reverse
-- |
-- like the chomp from Perl
-- but this chomp removes trailing newlines AND trailing spaces if any
chomp :: String -> String
chomp = reverse . chomp' . reverse
where chomp' [] = []
chomp' xs@(x:xs') | x == '\n' || x == ' ' = chomp' xs'
| otherwise = xs
-- IgnoreMaybe datatype
-- extension to Maybe for use in computations over recursive types
-- that need a "don't care" result
--
-- RealJust a means result a was computed
-- RealNothing means computation failed
-- IgnoreNothing means the rest of the computation should not be
-- influenced by this result
--
data IgnoreMaybe a = RealJust a
| RealNothing
| IgnoreNothing
-- drop IgnoreNothing from a list to get Maybe list
--
dropIgnore :: [IgnoreMaybe a] -> [Maybe a]
dropIgnore [] = []
dropIgnore ((RealJust x):t) = (Just x):(dropIgnore t)
dropIgnore (RealNothing:t) = Nothing:(dropIgnore t)
dropIgnore (IgnoreNothing:t) = dropIgnore t
-- convert from Maybe to IgnoreMaybe
--
toIgnore :: Maybe a -> IgnoreMaybe a
toIgnore (Just x) = RealJust x
toIgnore _ = RealNothing
-- convert from IgnoreMaybe to Maybe
-- IgnoreNothing is propagated (wrt to the meaning given above) to Nothing
--
toMaybe :: IgnoreMaybe a -> Maybe a
toMaybe (RealJust x) = Just x
toMaybe _ = Nothing
-- map over IgnoreMaybe taking (Maybe a -> b) function
--
mapIgnore :: (Maybe a -> b) -> [IgnoreMaybe a] -> [b]
mapIgnore f [] = []
mapIgnore f (IgnoreNothing:t) = mapIgnore f t
mapIgnore f (h:t) = (f $ toMaybe h):(mapIgnore f t)
-- map over IgnoreMaybe taking (a -> b) function
--
mapIgnoreMaybe :: (a -> b) -> [IgnoreMaybe a] -> [b]
mapIgnoreMaybe f [] = []
mapIgnoreMaybe f (RealJust x:t) = (f x):(mapIgnoreMaybe f t)
mapIgnoreMaybe f (h:t) = mapIgnoreMaybe f t
-- add element to list as if it were a set
-- addition is to the front if element wasn't already in the list, in which
-- case the list is not modified
--
setAddOne :: Eq a => [a] -> a -> [a]
setAddOne set x = if (x `elem` set) then set else (x:set)
-- concat two lists as if they were sets
-- adds each element from the second list using setAddOne
--
setAdd :: Eq a => [a] -> [a] -> [a]
setAdd set add = foldl setAddOne set add
-- find out whether all the elements of a list occur only once
--
allUnique :: Eq a => [a] -> Bool
allUnique [] = False
allUnique [h] = True
allUnique (h:t) = ([ x | x<-t, x == h ]==[]) && allUnique t
-- compute members of a list occuring more than once
--
notUnique :: Eq a => [a] -> [a]
notUnique [] = []
notUnique (h:t) = let
diff = [ x | x<-t, x/=h ]
rest = notUnique diff
in
case (h `elem` t) of True -> h : rest;
False -> rest